Impactites from the Hiawatha crater, North-West Greenland

(1)

Degree Project in Geology 60 hp

Master Thesis

Stockholm 2020

Department of Geological Sciences Stockholm University

SE-106 91 Stockholm

Impactites from the Hiawatha crater, North-West Greenland

Jacob Gustafsson

(2)

Abstract

The recent discovery of the 31-km-wide Hiawatha impact crater has raised unanswered questions about its age, impactor and highly unusual organic carbon component. Previous research suggests a fractionated iron meteorite impactor, a probable maximum 3–2.4 Ma impact age and a possible Younger Dryas impact age. The first objective in this study has been to investigate a possible link between the Cape York meteorites and the Hiawatha impact crater by comparing the chromium isotopic signature in chromite from a Cape York meteorite with the chromium isotopic signature in potential chromite from the Hiawatha impactor. The second objective has been to investigate a possible Hiawatha signature in the Younger Dryas deposits from Baffin Bay. The third objective has been to study the organic carbon component in impactites derived from the Hiawatha impact crater.

Heavy mineral grains were separated from glaciofluvial sediment which contains Hiawatha impactite grains. Not a single chromite grain was found and the possible link to the Cape York meteorites could not be tested. The petrographic examination of Younger Dryas marine deposits resulted in absence of impact-related Hiawatha grains. A petrological investigation revealed that organic carbon was likely found in five of six variably shocked impactites derived from the Hiawatha impact crater. The character of the organic carbon varies between the samples and also within individual samples. Vitrinite reflectance measurements of the organic carbon in two impactites yielded low reflectance values compared to charcoalification experiments of wood.

Organic particles with different reflectance in the same sample suggest that the particles had different impact histories prior to settling and becoming a rock. Diagnostic conifer cellular texture was found in at least one of the samples. The character of the organic particles in the impactites supports the suggestion in a previous study that the sources of the Hiawatha organic carbon component are unmetamorphosed surficial deposits containing dead conifer tree trunks and fine-grained layered clay and organic matter.

In this study it is concluded that the apparent absence of chromite in the examined glaciofluvial sediment sample corroborates the significance of previous research which suggests that the Hiawatha impactor wasan iron meteorite. The apparent absence of impact related grains in the Younger Dryas deposits suggests that although a Younger Dryas age for the Hiawatha impact crater is less likely now, the possibility remains open. The organic carbon with diagnostic conifer cellular texture in the Hiawatha impactites corroborates the conclusion in a previous study that the Hiawatha impact-related organic carbon component stems from local, thermally degraded conifer trees with a probable age of ca. 3–2.4 Ma. It is also concluded that the relatively low reflectance values of the organic carbon in the Hiawatha impactites seem to be related to the short duration of the high-temperature excursion during the hypervelocity impact event.

(3)

Abbreviations in microphotographs

BSE = Backscattered electron

EPI = Epi-illumination (reflected light) PPL =Plane polarized light

XPL = Crossed polarized light UV = Ultra violet

(4)

1. Introduction

The newly discovered 31-km wide Hiawatha impact crater (HIC) is highly unusual in many ways. It is one of the 25 largest impact structures on Earth (Kjær et al., 2018; Impact Earth, 2020, https:// impact.uwo.ca/). Even though it is probably the youngest known large impact structure on Earth (Garde et al., 2020), there is yet no proven ejecta layer associated with the HIC. It is the northernmost known impact crater on Earth and covered by the Greenland Ice Sheet (GrIS), which has an anomalous ice stratigraphy right over the crater compared to most of the GrIS (Kjær et al., 2018). The impact crater is located ~183 km from the proposed 36,5 km-wide Paterson crater, which like Hiawatha was discovered recently and is covered by the GrIS (MacGregor et al., 2019). The HIC is located ~300 km from Cape York, around where some of the largest known meteorites on Earth, the Cape York meteorites (CYM), were found.

The HIC has a highly unusual organic carbon component (Garde et al., 2020). All these circumstances have raised multiple questions of which the most important regards the age of the HIC.

Kjær et al (2018) reported the discovery of the impact crater and the quartz grains with shock-diagnostic planar deformation features (PDFs), which proved that Hiawatha is an impact crater. These grains were found in the sediment sample HW21-2016, which was deposited at the front of the Hiawatha Glacier just outside the crater rim. The sample contains a large variety of sand-sized impactite grains originating from the crater floor (Kjær et al., 2018). Elevated platinum (Pt) contents and anomalous platinum-group element elements (PGE) ratios in HW21- 2016 point to a strongly fractionated Duchesne (type IVA) iron meteorite as the Hiawatha impactor (Kjær et al., 2018). This chemical signature differs slightly from the CYM, which belongs to the less fractioned type IIIAB iron meteorites (Esbensen et al., 1982).

The following paragraph is based in information in (Kjær et al., 2018). Ice-penetrating radar evidence of the stratigraphy of the ice sheet over the HIC shows that the Holocene ice layer (11.7 to 0 ka) is continuous and conformable across the crater, whereas deeper and older ice is debris-rich or heavily disturbed. The debris clearly originates from the crater floor, indicating active subglacial erosion. The lack of distinct last glacial period (LGP; ~115 to 11.7 ka ago) reflections from pre-Younger Dryas (YD; ~12.8 to 11.7 ka ago) ice layers c. 100 km around the HIC indicate that the surrounding ice flow has also been disturbed. Moreover, radargrams of the ice layers shows evidence of subglacial water storage and active basal melting within and beneath the ice, which indicates that the ice is currently responding to a heat source.

Based on modern surface velocities (~10 to 30 m year−1) (figure 2D), it would only take a few millennia for deeper ice to flow across the crater, which excludes steady, uninterrupted ice flow from the ice sheet into the crater as a valid explanation for the ice stratigraphy over the crater.

Kjær et al (2018) interpreted the anomalous ice stratigraphy to be caused by a transient that strongly affected ice flow there after most of the LGP ice was deposited. Three hypotheses for such a transient were suggested: (i) The retreat of Humboldt Glacier around 9 to 8 ka, which unblocked the Nares Strait, (ii) Water pooling subglacially within the pre-existing crater and a subsequent catastrophic jökulhaup outburst (and possibly repeatedly) through the rim breach of the crater, or (iii)A change in ice flow as a response to the hypervelocity impact that formed the crater – if it occurred when ice was present there. The latter hypothesis suggests that the HIC could have been formed in the YD.

Garde et al (2020) investigated pebble-sized lignite lumps and gravel-sized charcoal particles, collected in the outwash sediments in front of the Hiawatha Glacier, together with carbonaceous impactite grains from sample HW21-2016. It was shown that the lignite lumps

(7)

2

largely consist of Pinus or Picea, with greatly expanded cork cells and desiccation cracks which indicate rapid, heat-induced expansion and contraction. Vitrinite reflectance (%Ro) of the charcoal pieces (Ro ≤ 3.5%) from sample HW21-2016, indicated impact-induced incineration of surficial and/or originally shallowly buried plant material, whereas the lignite (Ro ∼0.2%–

0.5%), presumably from the same hypervelocity impact, indicated a distinct heating episode but not direct incineration. Garde et al (2020) described a possible scenario in where the particles with lower Ro values could have been protected by the GrIS with associated permafrost in the target rocks, during the hypervelocity impact. By identifying the wood fragments to genera level, Garde et al (2020) concluded that the age of the modified carbon from the Hiawatha impact is probably 3–2.4 Ma, and thus also the probable maximum age of the impact. This conclusion was based on the fact that conifer trees, such as Pinus and Picea, are presently extinct from North Greenland, but known from late Pliocene deposits in the Canadian Arctic Archipelago (Fyles et al., 1991) and early Pleistocene deposits at Kap København (figure 2A) (Funder et al., 2001) in eastern North Greenland.

Even though the first firm evidence, in the form of thermally degraded conifer trees, provided of a probable maximum age (ca. 3-2.4 Ma) of the HIC, it is yet undated. It is also unknown what caused the anomalous ice stratigraphy over the crater. One of the previously described possible interpretations in Kjær et al (2018) implies that the HIC could be as young as the beginning of the YD. The evidence for an iron meteorite impactor for the Hiawatha impact, together with the close proximity to the finding places of the CYM, suggest a possible connection to the CYM. In order to address these research questions, 40 glacially transported pebbles and boulders (sample ID prefix = HW19) were collected at the ice margin adjacent to the HIC in the summer of 2019. A preliminary petrographic examination of the HW19 samples revealed that the majority of these samples are impactites and a high proportion of these impactites contain definite or suspected organic carbon.

There are three objectives in this thesis. The first objective is to investigate the possible link between the CYM and the HIC. If the CYM are related to the Hiawatha impact and if the terrestrial age of the CYM can be determined, we would have an indirect age determination for the HIC. Integrated with this investigation a new line of research is trying to absolute date the atmospheric entry of the CYM swarm. The second objective is to investigate if the HIC can be identified in YD deposits from Baffin Bay. If Hiawatha-related grains can be found in YD deposits, we would have an indirect age determination for the HIC. The third objective is to contribute to an ongoing assessment of the origin and nature of the highly unusual carbon component in the HIC. By investigating the organic carbon component in the HIC, we can get a better understanding of how organic matter behaves during a hypervelocity impact. The specific objectives in this thesis are the following:

• Find chromite in the glaciofluvial sand the sample HW21-2016 and compare chromium (Cr) isotope ratios in the chromite with Cr isotope ratios in chromite from the CYM.

• Search for ejecta and/or other impactite grains in YD deposits from Baffin Bay.

• Identify organic carbon in the HW19 impactites and distinguish it from e.g. iron oxides or hydroxides.

• Measure the reflectance (%Ro) of the organic carbon in the HW19 impactites.

(8)

3

2. Geological background

2.1 Impact structures in Greenland

Greenland, the largest island on Earth, is almost entirely covered by ice. As of today, the HIC is the only confirmed impact crater in Greenland (Impact Earth, 2020, https:// impact.uwo.ca/).

For comparison, there are 27 known impact craters located in Australia (figure 1). There are two other proposed impact structures on Greenland. Based upon multiple lines of evidence, Garde et al (2012), proposed that the 100 km scale Maniitsoq structure, in western Greenland, is the remnant of a giant c. 3 Ga hypervelocity impact. Based upon satellite and airborne geophysical data, MacGregor et al (2019), proposed an impact origin for a 36.5-km wide subglacial structure under ~2 km of ice in NW Greenland. The structure has an elevated circular rim as well as elevated central peaks, which could be the remnants of a central uplift. The minimum age of the ice overlying the structure is 79 ka (MacGregor et al., 2019).

Figure 1. 200 craters are currently plotted on the map as yellow pins (Impact Earth, 2020, https:// impact.uwo.ca/).

2.2 Sample locations

The key localities for the investigated material in this thesis are showed in figure 2A and B.

The HIC and the proposed Paterson crater are showed with red circles (figure 2A). The glacially transported pebbles and boulders (HW19) that were investigated in this thesis were sampled at the localities HW19-A, HW19-B and HW19-C (figure 2B). HW19-A is located in the foreland of the Hiawatha Glacier, which is protruding from the GrIS margin which is masking the HIC.

HW19-B and HW19-C are located at the curvilinear margin of the GrIS. Figure 2B also shows the sample location of the glaciofluvial sand sample HW21-2016. This sample site is located at the front of the Hiawatha Glacier just outside the crater rim. HW21-2016 was deposited no earlier than the year of 2010 (Kjær et al, 2018). The bulk mineral assemblage and chemical composition of the sand sample HW21-2016 indicate sourcing from local paragneiss (Garde et al., 2020). Based upon erratic boulders of paragneiss deposited by the Hiawatha Glacier, the target rock of the Hiawatha impact is assumed to be comprised of the same paleoproterozoic granulite-grade paragneiss exposed in the ice-free foreland to the crater, Inglefield Land (Kjær et al, 2018). Locally, this bedrock includes sulfidic paragneiss with graphite flakes (Dawes, 2004). In figure 2C, the paragneiss is represented by the pink Precambrian rock unit. Inglefield

(9)

4

Land is part of the E–W-trending Inglefield mobile belt which forms a 1.985 to 1.740 Ga old crystalline shield. (Dawes, 2004). Most of the CYM have been found in the vicinity of the Cape York peninsula and Meteorite Island (figure 2A). The marine sediment that was investigated in this thesis was extracted from site 012 (figure 2A).

Figure 2. A) Regional view of northwest Greenland. Modified from Kjær et al (2018). B) Elevation map of the bedrock showing the circular depression of the crater and sample locations. The ice margin of the GrIS is marked with the white line. Modified from Kjær et al (2018); Garde et al (2020). C) Bedrock type and lineations across Inglefield Land adjacent to the crater. The photograph shows a narrow zone along the ice margin where moderately to steeply dipping and commonly brecciated brittle planar structures (yellow lines indicate their locations and their strike) are superimposed on the Palaeoproterozoic bedrock foliation. The structures are tangential to the mostly subglacial rim of the HIC and dips away from it suggesting that they could be the outer, overturned part of the crater rim (Kjær et al (2018). D) A 5-m ArcticDEM mosaic over eastern Inglefield Land. The ice surface velocity is colour coded. The blue line illustrates an active basal drainage path inferred from radargrams of the ice From Kjær et al (2018).

A B

C D

(10)

5

2.3 The Cape York meteorites

The CY iron meteorites are fragments from one of the largest known meteorite showers on Earth (Appelt et al., 2015). The meteorites are named after the Cape York peninsula (figure 2A), around where some of the meteorites were found. At least ten specimens of the CYM, ranging from 290 g to 31 tons in weight (Buchwald, 1985), have been discovered (Appelt et al., 2015). The second largest CY meteorite, Agpalililk (20 tons), was discovered in the summer of 1963, by Dr. Vagn Buchwald. The main part (~15 tons) of Agpalilik is presently exhibited at the courtyard of the Geological Museum in Copenhagen. Agpalilik was found on an ice-free slope ~500 m from the shore at an altitude of 75 meters (Buchwald, 1975). Other CYM have also been found relatively close to the shore and at considerable altitudes (table 1). The meteorites have partly been covered with gneiss boulders and without any impact pits (Appelt et al., 2015). Although the CYM are essentially undated, Buchwald (1985) argued that the main body of the CYM probably entered the Earth’s atmosphere more than 2 ka, judging from the state of corrosion of the meteorites. Iron meteorites are divided into 13 subgroups based on their chemical composition. The CYM belong to the group IIIAB, which is the largest of 13 subgroups of iron meteorites. IIIAB contains about 300 known meteorites (Haack et al., 2006).

Unlike all of the other iron meteorites that are known (Appelt et al., 2015), the CYM are chemically diverse on a meter-scale (Esbensen et al., 1982; 1982a).

Table 1. Data of some of the Cape York meteorites. From Buchwald (1975).

2.4 Marine sediment from site 012

2.4.1 Core description and sediment age

In August 1991, three cores were retrieved in a trough close to the Greenland coast (site 012 in figure 2A), at a water depth of 823 m during the CSS‘Hudson’ cruise 91–039 at 76148.30N, 71151.50W. Site 012 had been identified as a favorable locality on a previous cruise. The sedimentary record from site 012 was subsequently investigated by Knudsen et al (2008) in a multiproxy study of the palaeoceanographic and paleoclimatic changes in northernmost Baffin Bay since the deglaciation of the area at about 12 500 calendar years before present (cal. yr BP).

Three cores were retrieved at site 012: the piston core 012P, the trigger weight core 012TWC and the box core 011BC. The sedimentary record from Site 012 were subdivided into eight time-intervals (12-1 to 12-8) based on major changes in environmental parameters (Knudsen et al., 2008).In this thesis, sediment from five different depth intervals from the piston core 012P were investigated: (i) 190-192 cm, (ii) 192-194 cm, (iii) 210-212 cm, (iv) 220-222 cm and (v) 230-232 cm. The ages in cal. yr BP (table 2) for these depth intervals, are based on calibrated accelerator mass spectrometry (AMS) ¹⁴C age determinations (Knudsen et al., 2008; see their

(11)

6

tables 3 and 4 for AMS ¹⁴C age determinations and an age model, respectively) performed at the Tandem Accelerator Laboratory in Uppsala, Sweden (Knudsen et al., 2008).

Table 2. Sediment intervals with ages in cal. yr BP. From Knudsen et al (2008).

2.4.2 Sediment lithology

Deglaciation of the area took place at about 12 500 cal. Yr BP, but the maximum expansion of the Weichselian ice south of Smith Sound (figure 2A) is unknown (Knudsen et al., 2008). The diamicton in the bottom of core 012P (figure 3) is assumed to be of late LGP age, although a Saalian or an even older age has been discussed (Kelly et al., 1999).

Interval 12-1

The sediment in interval 12-1 (205-240 cm) is comprised of a diamicton (Knudsen et al., 2008).

Blake et al (1996), suggested that the site was glaciated and that the sediments were deposited by glacial processes as a till. According to England et al (2006), the Nares Strait was filled by glaciers during the late LGP, and the basal diamicton at Site 012 might have been deposited during the deglaciation of the late LGP.

Interval 12-2

The sediment in interval 12-2 is comprised of laminated clay and silty clay without any microfossils and a negligible amount of organic carbon, probably indicating turbid waters and seasonally variable sedimentation from meltwater plumes during deglaciation (Knudsen et al., 2008)

Interval 12-3

The sediment in interval 12-3 is comprised of extremely fine laminated clay and silt-sized sediment in the lowermost laminated part of the interval. Foraminiferal assemblages appear within the laminated sediments at c. 12 300 cal. yr BP (Knudsen et al., 2008)

(12)

7

Figure 3. Lithological log and ¹⁴C age determinations for the studied intervals of cores 012P, 012 TWC and 011BC (Site 012). The vertical scale shows actual core barrel depths.

The column width is based on smoothed mean grain size with increasing grain size towards the right. From Knudsen et al (2008).

(13)

8

3. Theoretical background

3.1 Meteorite identification by Cr isotopes

Cr isotopes can be used to identify the meteorite type for an impact event (Koeberl et al., 2007).

First, the ⁵³Cr/⁵²Cr ratio is measured in the rock sample. The ⁵³Cr value is the daughter product of the extinct radionuclide ⁵³Mn (half-life = 3.7 Ma; Lugmair and Shukolyukov, 1998).

Secondly, the relative abundance of ⁵³Cr is measured as the deviation of the ⁵³Cr/⁵²Cr ratio in a sample relative to the standard terrestrial ⁵³Cr/⁵²Cr ratio (Koeberl et al., 2007). Such measurements are performed by high-precision thermal ionization mass spectrometry (TIMS) (Lugmair and Shukolyukov, 1998). All terrestrial ⁵³Mn was derived from the primordial solar nebula (Koeberl et al., 2007). Since the homogenization of the Earth was completed after all

53Mn on the planet had decayed, terrestrial rocks are not expected to show any variation in the

53Cr/⁵²Cr ratio. Most meteorite groups show a variable excess of ⁵³Cr relative to terrestrial samples (Koeberl et al., 2007).

3.2 Impactite formation

French (1998), defined impactites as, “all rocks affected by, or produced by, the shock waves and other processes generated by hypervelocity meteorite impact events” (p. 61). The pressures and associated shock metamorphism for a hypervelocity impact event are far beyond the range for normal geological processes (figure 4) (Stöffler et al., 2018). Under such extreme conditions shock waves will produce characteristic shock effects in minerals, such as mechanical deformation, phase transformation, decomposition and melting (Stöffler et al., 2018). The shock waves travel several km/s through rocks and affect their minerals with high and transient pressures, which results in shock-deformation effects of high stress, high strain rates (French, 1998). Along with the shock waves, energy is deposited in the minerals and therefore a particular shock pressure will generate a specific postshock temperature (figure 4). The postshock temperatures does not only increase along with pressure, but is also highly dependent on the target material. If the pressure and associated temperatures are high enough melting and vaporization can occur. What types of impactites are formed also depends on the target lithology, stratigraphy and the nature of the impactor and its impact angle (French, 1998).

According to Stöffler et al (2018), the formation of impactites depends on the following conditions:

1. The size and gravity of the target

2. The relative impact velocities of the impactor and target 3. The relative sizes of the impactor and target

4. The possible presence of an atmosphere and water on the target

5. The physical properties of the impactor and target (e.g., their density and porosity) A given impactor velocity will generate a variety of peak shock pressures and temperatures, depending on the porosity and shock impedance of a target. Shock impedance is defined as the product of shock wave velocity and the density of the mineral (Stöffler et al., 2018). Mechanical impedance determines how a shock wave reflects when it hits a boundary between materials with different densities and wave speeds (Ormö et al., 2013). At a given shock pressure, the shock temperature of highly porous rocks can be up to an order of magnitude higher (figure 4) than those of dense crystalline rocks (Stöffler et al., 2018). Compared to crystalline rocks, porous rocks have a higher capability to absorb the shock wave energy by the numerous grain

(14)

9

interfaces and pore spaces (French, 1998). No consensus has been reached so far, with respect to the onset or nature of specific shock effects in carbonate target rock (Stöffler et al., 2018).

Figure 4. Pressure–temperature fields for shock metamorphism of silicate rocks in comparison with endogenous metamorphism of the terrestrial crust. From Stöffler et al (2018).

3.3 Types of impactites

The following section is based upon French (1998). A variety of impactites are formed during an impact event. The sketch in figure 5 shows the basic distinction of impactites.

Parautochthonous rocks are beneath the crater floor and the allochthonous rocks are either crater-fill rocks or ejecta which is deposited outside of the crater. The parautochthonous rocks are subjected to relatively low shock pressures. Shock pressures for parautochthonous rocks may be up to 25-30 GPa in the center of the impact structure and may be less than 2 GPa at the rim. The shock pressures also decrease with depth and hence most of the parautochtonous rocks are subjected to less than 2 GPa. The shock effects for these impactites are mostly fracturing, brecciation and the formation of shatter cones. The allochtonous rocks comprises both the rocks that fill the crater after its excavation and also ejected material beyond the crater rim. These rocks are mainly subjected to relatively high shock pressures. The allochtonous rocks that fill the crater are called crater-fill breccias and are generally a complex mixture of impact melts, highly shocked, excavated and reposited rocks and also unshocked rocks from the crater rim and walls. Excavated rocks may be subjected to relatively high shock pressures of about 5 GPa to >100 GPa. More than 90% of the ejected material beyond the crater rim is deposited within five crater radii. This material is called proximal ejecta. Material that has been ejected beyond five crater radii is called distal ejecta. Distal ejecta tends to display a high level of shock effects.

Smaller ejecta particles (typically ≤1 mm) can be distributed by atmospheric transport to

(15)

10

regional or global distances. Ejecta deposits may contain shocked rock and mineral fragments, distinctive chemical and isotopic signatures and glassy objects such as tektites (French, 1998).

Figure 5. Cross section of a complex crater and the typical locations for proximal impactites. A complex crater is characterized by a centrally uplifted region, a generally flat floor and extensive inward collapse around the rim (French, 1998). On Earth, the transition between smaller bowl-shaped simple craters and complex craters occurs at a diameter of about 4 km in massive crystalline rocks and at about 2 km in sediments (French, 1998). Lithic breccias are melt-free impact breccias. Suevitic breccias are melt-bearing impact breccias Stöffler et al (2018).

Modified from Stöffler et al (2018).

3.4 Organic matter in impactites

Organic matter is an uncommon component in impact structures and occurs as recycled carbon from target rocks (Garde et al., 2020). Organic carbon in ejecta derived from sedimentary target rocks, has been investigated from the ca. 23 Ma Haughton crater in Canada (Parnell et al., 2007), as well as from the ca. 15 Ma Ries crater in Germany (Osinski, 2003). Charcoal particles in a drill core from the 66 Ma Chicxulub crater in Mexico, were investigated by Gulick et al (2019).

In some studies, the source of the carbon has been identified and used to constrain the maximum age of the impacts. Gilmour et al (2003) and Parnell and Lindgren (2006), argued that dispersed graphitic carbon in the ca. 500 Ma Gardnos crater in Norway stems from Cambrian Alum Shale. Howard et al (2013) described preserved biomarkers from vegetation in impact glasses, which were linked to the proposed ca. 0.8 Ma Darwin crater, Tasmania. Losiak et al (2016) investigated locally derived charcoal in the small ca. 3.2 ka Kaali craters, Estonia.

As previously described, Garde et al (2020) investigated thermally degraded organic material which yielded a probable maximum age for the Hiawatha impact of ca. 2.4-3 Ma.

Organic matter has also been found in impact glass unlinked to any known impact craters. Shultz et al (2014), investigated plant matter encapsulated within impact glasses found in different layers of the Pampean strata of Argentina. The impact glasses were produced by separate bolide impacts into the loessoid sediments between 9.2 Ma and 6 ka (Shultz et al., 2014). Moore et al (2020), described a 12.8 ka YD boundary layer at Abu Hureyra in Syria, which contains melt glass most likely from an airburst event, possibly accompanied by ground impacts by fragments of an impactor. Approximately 40% of the melt glass contains carbon- infused, siliceous plant imprints formed at a minimum of 1200°–1300°C (Moore et al., 2020).

(16)

11

4. Methods

4.1 Heavy mineral identification

4.1.1 Wilfley table gravity separation

Heavy mineral grains from the sand-sized (<600µm) glaciofluvial sample HW21-2016, were concentrated through a Wilfley table (Holman-Wilfley Ltd) gravity separation. Prior to the gravity separations the weight of HW21-2016 was 295 grams. The front tilt of the Wilfley table (figure 6A) was set to 7.5 degrees and the side tilt of the table to 1.6 degrees. The heaviest mineral grains in both sand samples were concentrated into three different fractions, which are called the “head”, the “tail” and the “extra tail”. Figure 6B shows the heaviest mineral grains remaining towards the end of the gravity separation. Finally, all sand that is left on the table is a thin line with the heaviest mineral grains from the uppermost groove. The lowermost half of this thin line is the tail fraction and the uppermost half is the “head" fraction.

Figure 6. A) Wilfley table separation of one of the sand samples. The table is tilted in two directions B) Only the heaviest sand is left towards the end of the Wilfley table separation.

4.1.2 Frantz magnetic separation and heavy liquid separation

Another subsample of HW21-2016 was separated with a Frantz isodynamic magnetic separator and with heavy liquids by Geotrack International Ltd in Australia. The heavy liquid separation was conducted at the separation specific gravities (SG) of 2.95 and 3.3. The subsample of HW21-2016 was separated into four fractions:

1) Vertical Frantz magnetics – grains removed before the heavy liquid separation 2) Horizontal Frantz magnetics – magnetic grains removed from the >2.95 SG heavy

mineral fraction

3) Apatite fraction – non magnetic 2.95 – 3.3 SG grains 4) Zircon fraction – non magnetic > 3.3 SG grains 4.1.3 Hand picking, magnetic separation and mounting

Three separates, the HW21-2016 “head” fraction, the HW21-2016 horizontal Frantz magnetics and the HW21-2016 zircon fraction, were selected for identification and placed in petri dishes filled with ethanol. In the HW21-2016 “head” fraction and the HW21-2016 horizontal Frantz magnetics, ferromagnetic grains were separated from weaker magnetic grains by a handheld magnet and a stronger neodymium magnet (figure 7A). In addition, grains were placed along the edge of a medium-sized petri dish which was moved around in a circular motion, so that the heavier grains were concentrated closer to the edge of the petri dish. A total of 123 grains were handpicked from the petri dishes with tweezers and pipettes under a stereomicroscope (Wild Heerbrugg). The selection was mainly based on the color, luster and shape of the grains.

Finally, the selected grains were mounted on epoxy glue, which together with epoxy resin and a hardener were made into a grain mount (figure 7B). In preparation for the subsequent micro-

A B

(17)

12

analysis with Automated Quantitative Mineralogy (AQM), the grain mount was polished with sand papers. The grains were placed with their flattest surface on top of the epoxy glue to minimize the mass loss by polishing.

Figure 7. A) Magnetic grains separated from nonmagnetic grains with a neodymium magnet. B) The final polished grain mount with 123 grains in eight rows.

4.1.4 Automated Quantitative Mineralogy

AQM is an integrated system with an energy dispersive spectrometry (EDX) detector mounted to a SEM integrated with software. The AQM system was developed in the 1980s to analyze the mineralogy, chemistry, and microstructures of mineral ores, fly ashes, and sediments (Keulen et al, 2020). The previously described polished grain mount was carbon coated and analyzed by AQM at the Geological Survey of Denmark and Greenland’s (GEUS) in Copenhagen. This AQM system includes a ZEISS SIGMA 300VP SEM equipped with a BSE detector and two Bruker XFlash 6j30 EDX detectors, with 129 eV energy resolution and with the ZEISS Mineralogic automated quantitative mineralogy software platform. The ZEISS Mineralogic software has integrated matrix corrections and full quantification of energy dispersive spectrometry data and is therefore able to give detailed chemical information on each pixel in the AQM mineral maps (Keulen et al, 2020). In this study AQM was applied for mineral maps and BSE images of the grains in the grain mount.

4.2 Marine sediment identification

4.2.1 Thin section preparation

The mineral identification was performed on lithogenous marine sediment from the marine piston core 91-039-012P. A total of 389 sand grains (figure 8A) were handpicked with a tweezer under a stereomicroscope of the model Olympus SZ40. The grains were sorted into sub-groups based upon their core depth. Impactite grains from the sample HW19 AB3 2019, which was collected in the foreland of the Hiawatha Glacier, were used as reference grains. Figure 8B shows some of the grains during the manufacturing of one of three grains mounts.

A B

(18)

13

Figure 8. A). Two of the 389 selected sand grains. B) Marine grains placed on top of yellow glue.

4.2.2 Optical mineral identification

The three grain mounts were sent to Precision Petrographics Ltd. in Canada, where three polished thin sections (figure 9) were made. The thin sections were examined with petrographic microscopes (Leica DMLP and Leica LEITZ DMRXP) with reflected and transmitted light illumination. Both plane polarized and crossed polarized light were used. Some of the grains were analyzed in the same SEM-EDS session as the previously described HW19-thin sections.

Figure 9. Three polished thin sections with the selected grains from the marine piston core 91-039-012P.

4.3 Carbon petrology

4.3.1 Thin section preparation

A total of 72 rock chips were cut from the 40 previously described rock HW19-samples (sampling locations in figure 2B) at the University of Copenhagen as the first step of thin section preparation. In figures 10A and B, rock samples HW19-39 and HW19-32 are shown. The rock chips were subsequently sent to Precision Petrographics Ltd, Canada, where 72 thin sections were manufactured. The thin sections were polished in order to allow observations in reflected light and non-destructive scanning electron microscope imaging and chemical analysis. 15 thin sections from as many rock chips (appendix A1) were selected for this study.

A B

(19)

14

Figure 10. A) Cross section of sample HW19-39. Arrow shows black piece of organic carbon. The diameter of coin is 2.25 cm. B) Cross section of sample HW19-32. Ruler (cm) for scale.

4.3.2 Petrographic microscopy

The thin sections were studied by petrographic microscopy, which was the main tool for mineral identification and establishing textural relationships. The main petrographic microscopes used in this study were a Leica DMLP and a Leica LEITZ DMRXP. The samples were examined in reflected and transmitted light. Both plane polarized and crossed polarized light were used. Out of the 15 thin sections, six were selected for further analyses with scanning electron microscopy and vitrinite reflectance.

4.3.3 Scanning Electron Microscope - Energy Dispersive Spectroscopy

Six thin sections were analyzed with a FEI Quanta 650 FEG Scanning Electron Microscope (SEM) at the Swedish Museum of Natural History in Stockholm, Sweden. The SEM was equipped with Energy-Dispersive Spectroscopy (EDS), which was used in order to obtain element concentrations from a point or the average from an area of the thin sections. Two different EDS detectors were used – an Oxford Xmax^N 80mm² SDD and an Oxford X Xmax^N 50mm² SDD. The accelerating voltage of the electron beam was set to 20 kV at a working distance of 10 mm. The SEM was operated in low vacuum mode (40 pascal), in where a micro atmosphere of water vapor functions as an electrically conductive medium, instead of using a traditional coating of carbon or gold. Copper tape or graphite tape were attached to the edges of the thin sections for enhanced conductivity. Carbon coating was avoided because carbon was an analytical target. Backscattered electron (BSE) imaging was used to reveal microstructures and mineral textures at both low and high magnification. The BSE detector shows the variations in mean atomic weight of the sample surface. The software AZtec (Oxford Instruments) was used for data reduction. Scanned close-ups of the thin sections were printed out in grayscale on A4 paper sheets and used as orientation maps during the SEM-EDS sessions.

4.3.4 Vitrinite reflectance

Six thin sections were analyzed quantitively with vitrinite reflectance (VR) at the Lithospheric Organic Carbon Laboratory (L.O.C.) laboratory in Aarhus University, Denmark. VR data is typically presented in units of reflectance (%Ro), which is the measured percentage of reflected light from a sample immersed in oil (Burnham and Sweeney, 1989). In this study, carbon petrology was carried out using a Zeiss Imager M2m AXIO microscope, which was equipped with 50x oil immersion and 63x water immersion objectives and the Diskus-Fossil program (Hilgers Technisches Buero,Konigswinter, Germany). Random reflectance measurements (%Ro) were conducted under oil immersion using the ×50 objective on assumed carbonaceous matter in incident white light and the microscope system was calibrated against a KB N-LASF 1.32%Ro standard.

A B

(20)

15

5. Results

5.1 Heavy mineral identification

Out of the 123 grains from the HW21-2016 sample that were identified by AQM, six grains were from the Wilfley table separated heavy mineral “head” fraction, 11 grains were from the horizontal Franz magnetic fraction (>2.95 SG) and 106 grains were from the non-magnetic zircon fraction (> 3.3 SG). Not as single grain was identified as chromite. In table 3, the bulk mineral assemblage and chemical composition are shown for all grains from the “head” fraction (row 1), for all grains from the horizontal Franz magnetic fraction (row 2) and for 26 grains from the zircon fraction (row 3).

The mineral map in figure 11 and the raw data for row1 in table 3, reveal that ilmenite and Fe-oxide are the main mineral components of the grains in the “head” fraction. Most of these grains also have a minor rutile component. The grains are dark-coloured, opaque and have a dull or submetallic luster. The shapes of these grains are mainly equant. The two grains with a high proportion of iron oxide are more tabular (figure 11).

In table 3, the 11 grains in row 2 are from the horizontal Franz magnetic fraction. The main mineral component in eight of the grains is ilmenite. These grains also have minor proportions of iron oxide and rutile. The main mineral component in the three remaining grains is iron oxide, pyroxene and rutile, respectively. The grains in row 2 are dark-coloured, opaque and have a dull or submetallic luster. The shapes of these grains are either equant, tabular or irregular.

The main mineral component in the remaining 106 grains is rutile. The raw data for the 26 grains in row 3 (table 3) is also representative for the remaining grains from the nonmagnetic zircon fraction. Mineral maps and raw data revealed that these grains have a relatively high proportion of an unclassified mineral component. This is also the case for the rest of the grains from the zircon fraction. The AQM data showed that this unclassified component has a relatively high proportion of element vanadium (between 3.32 to 4.05%). For instance, the proportion of vanadium in the unclassified component of the grains in row 3, is 4,05% (table 3). The majority of the selected grains from the zircon fraction are dark-coloured and opaque under a stereomicroscope. A minor proportion of the grains are dark brown-coloured and slightly translucent. The selected grains from the zircon fraction can be divided into two groups with respect to luster. One group has a duller luster and the other group has a more adamantine to submetallic luster. The most common grain shapes are equant or tabular.

(21)

16

Table 3. AQM data for three of the eight rows of grains in the grain mount (figure 7B) with selected grains from the HW21-2016 sample. Row 1 data are from the “head” fraction. Row 2 data are from the horizontal Franz magnetic fraction. Row 3 data are from the zircon fraction.

Row Mineral Area % Weight % Average Composition %

Row 1 Ilmenite 53,46 81,60 Fe 39,32; Ti 36,84; O 23,84;

Titanite 0,27 0,30 O 34,22; Ti 28,46; Ca 23,32; Si 14;

Rutile 4,26 6,09 Ti 62,43; O 37,57;

Biotite 0,01 0,01 O 33,19; Fe 29,15; Si 19,94; Al 10,2; K 7,52;

Fe-Oxide 37,74 12,00 Fe 64,21; O 23,2; Ti 12,47; Si 0,11; S 0;

Unclassified

4,27

-

Ti 37,27; O 30,26; Fe 20,1; Ca 5,41; Si 4,11; V 1,23; Al 0,92;

Mg 0,6; Na 0,05; Pb 0,02; As 0,01;

Row 2 Fe-Oxide 6,94 1,73 Fe 59,4; O 26,23; Ti 14,33; Si 0,03;

Chamosite 0,47 0,35 O 36,34; Fe 27,85; Si 21,75; Al 7,62; Mg 6,12; K 0,32;

Amphibole 0,39 0,30 O 44,87; Si 21,47; Fe 12,69; Ca 7,75; Al 6,73; Mg 6,49;

Clinochlore 0,01 0,00 O 41,21; Si 21,2; Fe 20,33; Mg 9,12; Al 8,15;

Pyroxene

13,44

11,02

O 31,87; Si 21,73; Fe 21,41; Ca 11,15; Al 6,96; Mg 5,93; Na 0,95; Zn 0,01; K 0,01;

Biotite 4,79 3,57 O 36,18; Si 23,96; Fe 23,08; Al 7,72; Mg 6,56; K 2,52;

Titanite 0,01 0,01 O 34,17; Ti 30,73; Ca 22,79; Si 12,31;

Ilmenite 64,00 76,33 Ti 38,58; Fe 37,51; O 23,92;

Rutile 5,99 6,69 Ti 62; O 38;

Unclassified

3,96

-

Ti 39,99; O 34,03; Fe 21,64; Si 1,33; V 1,25; Ca 0,73; Al 0,56; Mg 0,3; Na 0,06; K 0,04; Mn 0,03; Zn 0,01;

Ba 0,01;

Row 3 Rutile 78,92 99,96 Ti 68,81; O 31,19;

Biotite 0,01 0,01 O 51,09; Si 17,2; Al 15,15; K 7,3; Fe 5,86; Mg 3,41;

Kaolinite 0,01 0,01 O 51,96; Al 25,17; Si 22,87;

Ilmenite 0,02 0,03 Ti 40,57; Fe 31,31; O 28,12;

Unclassified

21,04

-

Ti 80,29; O 15,15; V 4,05; Fe 0,13; Si 0,08; Al 0,07;

Zr 0,06; Nb 0,04; Mg 0,04; Na 0,03; P 0,01; Ni 0,01;

Y 0,01; S 0,01;

Figure 11. The upper row shows BSE images of the six grains from the Wilfley table separated “head” fraction (Row 1 in table 3). The bottom row is a mineral map of the same grains.

(22)

17

5.2 Marine sediment identification

A total of 389 marine sand-sized grains were selected under a stereomicroscope for subsequent petrographic examination. The strategy was to search for ejecta or impactite grains, such as impact glass, microbreccia or quartz with planar deformation features (PDFs). Not a single grain in any of the five depth intervals contain any known shock metamorphic evidence. The only grain which was found to contain glass is the volcanic grain in figure 13C. In a few grains, linear features were found in quartz and feldspars. These linear features are not considered to be PDFs, but rather inclusion trails, rutile growth or cleavage planes. Some unclassified grains were found in the 230-232 cm and 210-212 cm samples.

Core depth 230-232 cm (older than 12 500 cal. yr BP)

Table 4 shows a general description of the 230-232 cm sample under a stereomicroscope, before any grains were examined under a petrographic microscope. The grains that were considered to be of the most interest was dark-coloured grains with inclusions and polymineralic crystalline grains.

Table 4. Types of grains and their relative quantity in the 230-232 cm sample.

Relative quantity Types of grains

Abundant • Monocrystalline quartz grains.

• Composite grains of mica and quartz.

• Dark-coloured, rounded grains with inclusions. Some of these grains are relatively soft.

Moderately abundant • Sedimentary composite grains of felsic composition. These grains are easily broken apart and held together by a fine-grained cement.

• Felsic polymineralic crystalline grains.

• Mafic polymineralic crystalline grains

Rare • Dark-coloured monocrystalline grains with submetallic luster.

• Other grains.

The subsequent petrographic examination revealed that most of the examined dark-coloured grains are fine-grained sedimentary grains. The mafic polymineralic crystalline grains are mainly classified as volcanic grains. The felsic polymineralic crystalline grains were unclassified. Figures 12A and B show a grain or probable volcanic origin with brown-coloured matrix and mineral fragments with unordered orientations. Some of the fragments are probable mica fragments. Figures 12C-D shows two unclassified grains. Minerals within the matrix have concave boundaries. In figure 12C a deformed mineral fragment is showed. Not a single grain in this sample showed any evidence for shock metamorphism.

(23)

18

Figure 12. A) Grain with brown-coloured matrix of probable volcanic origin. B) BSE image of the same grains as in A) Arrow show slightly bent probable mica. C) Unclassified grain. Arrow shows a deformed mineral, probably mica. D) Unclassified grain. Arrow shows a mineral with concave boundaries.

Preliminary microscopy revealed that the grain types in the 220-222 cm sample are mostly similar to the grain types in the 230-232 cm sample (table 4). In general, the grains in the 220- 222 cm sample are slightly larger compared to the grains in the 230-232 cm sample. The main difference is that the mafic polymineralic crystalline grains are more common in the 220-220 cm sample.

The subsequent petrographic examination revealed that not a single grain in this sample showed any evidence for shock metamorphism. The relatively common mafic polymineralic crystalline grains are classified as volcanic grains.

Preliminary microscopy revealed that the grain types in the 210-212 cm sample are generally similar to the grain types in the 230-232 cm sample (table 4). As for the 220-222 cm sample, the average grain size is slightly larger compared to the grains in the 230-232 cm sample and the mafic polymineralic crystalline grains are more abundant.

The petrographic examination revealed that the mafic polymineralic crystalline grains are volcanic grains with lathwork texture. One of these grains is shown in figures 13A and B.

It has a typical volcanic lathwork texture with plagioclase laths together with pyroxene. Figure 13B shows cleavage in two of the adjacent pyroxene crystals. Another volcanic grain with plagioclase laths, pyroxene and zoned secondary chlorite is shown in figures 13C and D. In figure 13E, an unclassified grain is shown. The bright rectangular crystals are plagioclase. The

PPL BSE

XPL

A B

C D XPL

(24)

19

material in the thin grayish stripe at the upper right grain boundary in figure 13E is quartz. The large opaque mineral has a unknown chemical composition. Its high reflectance in reflected light indicate an oxide or sulfide. Another unclassified grain is shown in figure 13F. It contains biotite fragments with opaque Fe-oxide boundaries. The biotite fragments are surrounded by chlorite. Not a single grain in this sample showed any evidence for shock metamorphism.

Figure 13. A) Volcanic grain with lathwork texture. B) Same grains as in A). Pyroxene with cleavage. C) Volcanic glass in zeolite. Plane polarized light D) Same grains as in C) in crossed polarized light. E) Unclassified grain.

White arrow shows plagioclase. Black arrow shows quartz. F) Unclassified grain. White arrow shows Fe-rich biotite boundaries.

B BSE

C D A

E F

XPL

PPL XPL

PPL PPL

100 µm 100 µm

(25)

20

Core depth 192-194 cm (younger than 12 300 cal. yr BP)

Preliminary microscopy revealed that the average grains size of the 192-194 cm sample is smaller compared to the average grains size for the latter three intervals and the proportion of quartz grains are significantly higher. No volcanic grains were found in this sample. A high proportion of fine-grained consolidated sedimentary grains were found. Such grains were not observed in the three previously described samples. The subsequent petrographic examination of selected grains from this sample showed that none of these grains are impact related.

Core depth 190-192 cm (younger than 12 300 cal. yr BP)

Preliminary microscopy revealed that the grain types in the 190-192 cm sample are mostly similar to the grain types in the 190-192 cm sample. The main difference is that the proportion of fine-grained consolidated sedimentary grains are higher in the 190-192 cm sample than in the 192-194 cm sample. The petrographic examination of the selected grains from this sample showed that none of these grains are impact related.

5.3 Carbon petrology of HW19 impactites

After a preliminary petrographic examination of 15 thin sections, six of these were selected for further examination based upon: (i) assumed organic carbon content, and (ii) degree of shock metamorphism. The six samples (figure 14) were subdivided into three categories based upon their estimated level of shock metamorphism: (1) weakly shocked, (2) moderately shocked, and (3) strongly shocked. Based upon the proposed shock classification systems for felsic and mafic rocks in Stöffler et al (2018), all six samples have diagnostic evidence for shock metamorphism.

It is however important to point out that shock-metamorphic features which provide definite evidence for meteorite impact origin (French, 1998), such as PDFs, were found in three of the samples (table 5). The other three thin section have diagnostic evidence for shock metamorphism but not evidence which can be considered to be diagnostic in a definite sense according to (French, 1998). It is also important to point out that one of the samples, HW19- 39d, contains a high proportion of carbonate. As previously described, the degree of shock metamorphism at a given shock pressure, depends primarily on the mineralogical composition and the porosity of a rock or sediment sample. Furthermore, no consensus has been reached so far regarding the onset or even nature of specific shock effects in carbonates (Stöffler et al., 2018). HW19-39d is classified as moderately shocked based upon evidence for melted carbonate. SEM-EDS measurements on assumed carbonaceous material, showed that five of the six samples clearly contain carbonaceous material. Table 5 summarizes the analyses that were performed and the subdivisions of the samples. In figure 14, the selected samples are sorted by their level of shock metamorphism which increases from the top row towards the bottom row.

(26)

21 Table 5. Sample summary. *Vitrinite reflectance (%Ro).

Sample Microscopy SEM-EDS VR* Shock level Diagnostic? Carbonaceous

HW19-02a Yes Yes Yes High PDFs in Qz Yes

HW19-07a Yes Yes Yes Low Brecciation Yes

HW19-23a Yes Yes Yes Moderate PDFs in Qz Yes

HW19-32a Yes Yes Yes High PDFs in Qz Yes

HW19-39d Yes Yes Yes Moderate Melted CaCO3 Yes

HW19-40a Yes Yes Yes Low Brecciation No

Figure 14. Overview images of the six selected HW19 samples. The samples on the top row are classified as weakly shocked impactites, the samples in the middle row are classified as moderately shocked impactites and the samples in the bottom row are classified as highly shocked impactites.

(27)

22 5.3.1 Weakly shocked impactites

HW19-07a – impactite lithology

Sample HW19-07a is classified as a microbreccia which contains brecciated silicate rock fragments of varying sizes. In the central part of the sample overview image (figure 14), relatively large rock fragments are cut by a relatively large, fine-grained and maroon-coloured vein, which consists of fine-grained albite and Fe-oxide. Relatively small brecciated rock and mineral fragments are located in the vein adjacent to its margin to larger rock fragments. For instance, figure 15A shows a relatively large opaque mineral fragment, with varying Fe- and Ti content (shown by SEM-EDS analysis) located adjacent to silicate rock fragments. The main evidence for shock metamorphism is the brecciated silicate rock fragments and the fine-grained maroon-coloured vein, which contains tiny minerals fragments of similar chemical composition as adjacent silicate rock fragments. The sample also contain hydrothermally deposited Fe-oxide (figure 15B) and secondary chlorite, which indicate that the rock has experienced post-shock hydrothermal alteration.

Figure 15. A) A relatively large opaque mineral fragment adjacent to silicate rock fragments. B) Arrow shows hydrothermally deposited Fe-oxide in a microbreccia.

HW19-07a – carbonaceous matter

In sample HW19-07a opaque matter was examined by SEM-EDS, which revealed that most of the measured material was Fe-oxide and only a small proportion was carbonaceous. Vitrinite reflectance measurements were performed on opaque fragments, which were shown to be Fe- rich minerals by SEM-EDS measurements. Figure 16 A-C show that carbon was detected in what appears to be microfractures within quartz. Spectrum 585 and 586 revealed carbon adjacent to biotite in a microfracture. The bright minerals are Fe-oxide. Carbon was also detected associated with Fe-oxide along the cleavage planes of albitic felspar (appendix B1).

EDS spectum 542 shows that this material is Fe oxide associated with carbon. In general, there seems to be a minute amount of dispersed carbon in HW19-07a.

A PPL B PPL

(28)

23

Figure 16. Carbonaceous material adjacent to biotite and Fe-oxide in what appear to be microfractures within quartz A) Plane polarized light B) Crossed polarized light C). Backscattered electron image and SEM-EDS spectra.

HW19-40a –impactite lithology

Sample HW19-40a is classified as microbreccia which contains relatively large semi-layered fragments and tiny quartz fragments in a phyllosilicate-rich matrix. The semi-layered fragments are either comprised of Fe-oxide and phyllosilicate or iron oxides and quartz. These fragments seem to be replacements rather than rock fragments, probably due to post-impact hydrothermal activity (A. Garde, personal communication). The main evidence for shock metamorphism is brecciated angular fragments of quartz, which testifies to mineral fracturing.

HW19-40a – opaque matter

In sample HW19-40a all opaque matter which was suspected to be carbonaceous were shown by reflected light microscopy and SEM-EDS analysis to be Fe-oxide. One such examples is shown in figure 17A and B. HW19-40a is therefore classified as non-carbonaceous.

A B

C

PPL XPL

BSE

(29)

24

Figure 17. Opaque Fe-rich fragment surrounded by a phyllosilicate-rich matrix and brecciated quartz fragments.

A) In plane polarized light B) In reflected light.

5.3.2 Moderately shocked impactites HW19-23a – impactite lithology

Sample HW19-23a is classified as a microbreccia with abundant angular silicate fragments of varying sizes in a Fe-rich brown-coloured matrix. A finely dispersed Fe-rich opaque material is abundant in between silicate fragments in the fine-grained matrix. This material exhibits high reflectance and melt grain textures. The main evidences for shock metamorphism are quartz fragments with PDFs and ballen quartz (figure 18 A and B). Ballen quartz may be related to the impact-produced formation of fused silica glass or diaplectic quartz glass (French and Koeberl, 2010).

Figure 18. A) Likely two sets of crossing PDFs in quartz. B) Likely ballen quartz.

HW19-23a – carbonaceous matter

Sample HW19-23a, contains only a small carbon component. In figure 19A, spectrum 51 shows that carbon was detected in a particle which has a biotite-like chemistry excluding the carbon.

In figure 19B and C a similar particle is shown. Figure 19C shows that the particles exhibit relatively low reflectance. These types of particles have a similar apperance as the partice in figure 19A and are quite abundant in the sample. Vitrinite reflectance measurements were performed on opaque fragments in sample HW19-23a, which were shown to be Fe-rich minerals by SEM-EDS measurements. Spectrum 52 might be from a sample hole.

A PPL B EPI

A XPL B XPL

(30)

25

Figure 19. A) Backscattered electron image and SEM-EDS spectra of carbonaceous fragments. B) Possible carbonaceous fragment. C) The same fragments as in B) in crossed polarized light.

HW19-39d - impactite lithology

Sample HW19-39d is classified as a polymict microbreccia comprised of mainly carbonate- and silicate fragments together with a high proportion of organic carbon particles in a Ca-rich (EDS spectrum 265, appendix B2) matrix. The matrix is brown-coloured and contains tiny brecciated fragments. Besides the brecciation, the main proof of shock metamorphism in the sample are kinked biotite (figure 20A) and melt-texture in carbonate clasts (figure 20B).

Another thin section from the same rock sample, HW19-39a, contains a carbonate-melt clast with carbonate-filled vesicles (A. Garde, personal communication). Notably, the overview image (figure 14) of the sample shows a crude NW-SE trending layering of the mineral fragments.

PPL EPI

A

B C

(31)

26

Figure 20. A) Kink bands in a biotite fragment surrounded by brown-coloured matrix and tiny mineral fragments. B) Melt-texture in a carbonate fragment.

HW19-39d - Carbonaceous matter

Sample HW19-39d contains a high proportion of carbonaceous particles. Some of these particles have cellular textures. A relatively large carbonaceous particle in the sample exhibit even-sized and blocky cellular texture figure 21A). Figure 21B shows a transverse cross- sectioned organic carbon piece with cellular texture. The formed cells are variably filled with the matrix material, which is why the voids in the cell centres have different sizes. SEM-EDS- measurements of carbonaceous matter from sample HW19-39d is shown in appendix B2.

Figure 21. A) Relatively large particle with cellular texture shown by white arrows B) Organic carbon piece with regular rows of former cells.

Vitrinite reflectance measurements of carbonaceous particles in sample HW19-39d revealed two populations with respect to Ro values. The largest population has Rovalues between ~0.3 to 1% (figure 22A). The histogram shows that the majority of the 76 measurements have Ro values < 0.5 %. Two of these measurements are shown in figure 22C. The second population, which has Ro values between 1 and 2%, is smaller and accounts for almost 20 % of the measurement in the histogram in figure 22A. Almost 50 % of the measured Rovalues for this population is between ~1.5 to 2%. The reflectance measurements in figure 22B were solely performed on the particle in figure 22D.

A XPL B XPL

A PPL B EPI

(32)

27

Figure 22. A) Histogram of 76 vitrinite reflectance measurements on multiple particles. B) Histogram of 21 vitrinite reflectance measurements on the same organic carbon particle. C) Two measurements with Ro < 0.5%.

D) 10 measurements of a single organic carbon particle. The measurements in B) are from the same particle.

5.3.3 Highly shocked impactites HW19-02a – impactite lithology

Sample HW19-02a is classified as a microbreccia. The brown-coloured matrix is semi- crystalline and contains prismatic pyroxene microlites (Spectrum 464, appendix B3). At some places the brown matrix exhibits flow-texture adjacent to mineral fragments (figure 23A). A high proportion of the quartz fragment are shocked, ”toasted” quartz (figure 23B). At least one granular zircon was found in another thin section, HW19-02b, which is from the same sample (A. Garde, personal communication)

.

Figure 23. A) Flow-texture in the brown-coloured matrix below the relatively large dark-coloured mineral fragment in the bottom centre of the image B) Shocked, ”toasted”, quartz with at least one set of PDFs.

A B

C D

A PPL B PPL

Relative Frequency

Ro Ro

Relative Frequency

Impactites from the Hiawatha crater, North-West Greenland

Degree Project in Geology 60 hp

Master Thesis

Stockholm 2020

Impactites from the Hiawatha crater, North-West Greenland

Jacob Gustafsson

Abstract

Table of Contents

1. Introduction

2. Geological background

2.1 Impact structures in Greenland

2.2 Sample locations

2.3 The Cape York meteorites

2.4 Marine sediment from site 012

3. Theoretical background

3.1 Meteorite identification by Cr isotopes

3.2 Impactite formation

3.3 Types of impactites

3.4 Organic matter in impactites

4. Methods

4.1 Heavy mineral identification

4.2 Marine sediment identification

4.3 Carbon petrology

5. Results

5.1 Heavy mineral identification

5.2 Marine sediment identification

5.3 Carbon petrology of HW19 impactites

.